WO2019119966A1 - Procédé de traitement d'image de texte, dispositif, équipement et support d'informations - Google Patents

Procédé de traitement d'image de texte, dispositif, équipement et support d'informations Download PDF

Info

Publication number
WO2019119966A1
WO2019119966A1 PCT/CN2018/112093 CN2018112093W WO2019119966A1 WO 2019119966 A1 WO2019119966 A1 WO 2019119966A1 CN 2018112093 W CN2018112093 W CN 2018112093W WO 2019119966 A1 WO2019119966 A1 WO 2019119966A1
Authority
WO
WIPO (PCT)
Prior art keywords
image block
image
information
text
transformation
Prior art date
Application number
PCT/CN2018/112093
Other languages
English (en)
Chinese (zh)
Inventor
王权
梁鼎
钱晨
Original Assignee
北京市商汤科技开发有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京市商汤科技开发有限公司 filed Critical 北京市商汤科技开发有限公司
Priority to JP2020511273A priority Critical patent/JP6961802B2/ja
Publication of WO2019119966A1 publication Critical patent/WO2019119966A1/fr
Priority to US16/693,616 priority patent/US11275961B2/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/24—Aligning, centring, orientation detection or correction of the image
    • G06V10/242—Aligning, centring, orientation detection or correction of the image by image rotation, e.g. by 90 degrees
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/14—Image acquisition
    • G06V30/146—Aligning or centring of the image pick-up or image-field
    • G06V30/1463—Orientation detection or correction, e.g. rotation of multiples of 90 degrees
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/14—Image acquisition
    • G06V30/148—Segmentation of character regions
    • G06V30/158—Segmentation of character regions using character size, text spacings or pitch estimation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/18—Extraction of features or characteristics of the image
    • G06V30/1801—Detecting partial patterns, e.g. edges or contours, or configurations, e.g. loops, corners, strokes or intersections
    • G06V30/18019—Detecting partial patterns, e.g. edges or contours, or configurations, e.g. loops, corners, strokes or intersections by matching or filtering
    • G06V30/18038—Biologically-inspired filters, e.g. difference of Gaussians [DoG], Gabor filters
    • G06V30/18048—Biologically-inspired filters, e.g. difference of Gaussians [DoG], Gabor filters with interaction between the responses of different filters, e.g. cortical complex cells
    • G06V30/18057—Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/19—Recognition using electronic means
    • G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
    • G06V30/1916—Validation; Performance evaluation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/19—Recognition using electronic means
    • G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
    • G06V30/19173—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/40—Document-oriented image-based pattern recognition
    • G06V30/41—Analysis of document content
    • G06V30/413—Classification of content, e.g. text, photographs or tables
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition

Definitions

  • FIG. 2 is a schematic diagram of some embodiments of a neural network provided by an embodiment of the present application.
  • the neural network in the embodiments of the present disclosure can be implemented in various ways.
  • the neural network may be a convolutional neural network, and the present application does not limit the specific network structure of the convolutional neural network.
  • the convolutional neural network includes, but is not limited to, a convolutional layer, a non-linear Relu layer, a pooled layer, a fully connected layer, etc., and the more layers the convolutional neural network contains, the deeper the network.
  • the network structure of the convolutional neural network may use, but is not limited to, a network structure adopted by a neural network such as an ALexNet, a Deep Residual Network (ResNet), or a VGGnet (Visual Geometry Group Network). .
  • the perspective transformation matrix is determined based on the perspective transformation information of the image block.
  • the perspective transformation information of the image block may include the displacement amount of the vertex transformation, and the perspective transformation matrix is obtained according to the displacement amount, and the perspective transformation matrix may change the position of each point in the image.
  • the image block shape transformation information includes: perspective transformation information of the image block, image block rotation information indicating an integer multiple of 90 degrees, and indication information indicating whether an area occupied by the text in the image block reaches a predetermined requirement.
  • an implementation manner of performing morphological transformation processing on the character image to be processed according to the image block shape transformation information includes:
  • the text recognition in the S130 may include determining text content in the to-be-processed text image, or performing text detection on the to-be-processed text image, and determining text content in the to-be-processed text image based on the text detection result. This is not limited.
  • the morphological change label information of the image block samples in the training data includes perspective transformation label information of the image block samples, image block sample rotation label information indicating a 90-degree integer multiple, and characters in the image block samples.
  • the occupied area reaches the predetermined required annotation information (ie, the neural network includes the shared neural network, the first branch, the second branch, and the third branch)
  • the information is first labeled with a perspective transformation of the image block sample.
  • the shared neural network and the first branch of the neural network are supervised and learned; after the first branch training is completed, the shared neural network and the network parameters of the first branch (such as weights, etc.) are fixed and expressed by 90 degrees.
  • RAM 803 various programs and data required for the operation of the device can be stored.
  • the CPU 801, the ROM 802, and the RAM 803 are connected to each other through a bus 804.
  • ROM 802 is an optional module.
  • the RAM 803 stores executable instructions or writes executable instructions to the ROM 802 at runtime, the executable instructions causing the central processing unit 801 to perform the steps included in the object segmentation method described above.
  • An input/output (I/O) interface 805 is also coupled to bus 804.
  • the communication unit 812 may be integrated, or may be configured to have a plurality of sub-modules (for example, a plurality of IB network cards) and be respectively connected to the bus.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Molecular Biology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Software Systems (AREA)
  • Mathematical Physics (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Biodiversity & Conservation Biology (AREA)
  • Databases & Information Systems (AREA)
  • Medical Informatics (AREA)
  • Image Analysis (AREA)
  • Character Input (AREA)
  • Character Discrimination (AREA)
  • Editing Of Facsimile Originals (AREA)

Abstract

L'invention concerne un procédé, un dispositif, un équipement, un support d'informations et un programme informatique de traitement d'image de texte. Le procédé de traitement d'image de texte consiste principalement : à acquérir au moins un bloc d'image comprenant un texte d'une image comprenant un texte à traiter ; à acquérir des informations de transformation morphologique de bloc d'image du bloc d'image sur la base d'un réseau neuronal, les informations de transformation morphologique de bloc d'image étant utilisées pour transformer la direction du texte dans le bloc d'image en une direction prédéterminée ; à acquérir le réseau neuronal acquis par apprentissage à l'aide d'échantillons de blocs d'image présentant des informations d'annotation de transformation morphologique ; à effectuer un traitement de transformation morphologique, sur la base des informations de transformation morphologique de bloc d'image, par rapport à l'image de texte à traiter ; et à réaliser une reconnaissance de texte par rapport à l'image transformée morphologique comportant le texte à traiter.
PCT/CN2018/112093 2017-12-22 2018-10-26 Procédé de traitement d'image de texte, dispositif, équipement et support d'informations WO2019119966A1 (fr)

Priority Applications (2)

Application Number Priority Date Filing Date Title
JP2020511273A JP6961802B2 (ja) 2017-12-22 2018-10-26 文字画像処理方法、装置、機器及び記憶媒体
US16/693,616 US11275961B2 (en) 2017-12-22 2019-11-25 Character image processing method and apparatus, device, and storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711407535.0 2017-12-22
CN201711407535.0A CN108229470B (zh) 2017-12-22 2017-12-22 文字图像处理方法、装置、设备及存储介质

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US16/693,616 Continuation US11275961B2 (en) 2017-12-22 2019-11-25 Character image processing method and apparatus, device, and storage medium

Publications (1)

Publication Number Publication Date
WO2019119966A1 true WO2019119966A1 (fr) 2019-06-27

Family

ID=62648429

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/112093 WO2019119966A1 (fr) 2017-12-22 2018-10-26 Procédé de traitement d'image de texte, dispositif, équipement et support d'informations

Country Status (4)

Country Link
US (1) US11275961B2 (fr)
JP (1) JP6961802B2 (fr)
CN (1) CN108229470B (fr)
WO (1) WO2019119966A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113326724A (zh) * 2021-02-07 2021-08-31 海南长光卫星信息技术有限公司 一种遥感影像变化检测方法、装置、设备及可读存储介质

Families Citing this family (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10198626B2 (en) * 2016-10-19 2019-02-05 Snap Inc. Neural networks for facial modeling
CN108229470B (zh) 2017-12-22 2022-04-01 北京市商汤科技开发有限公司 文字图像处理方法、装置、设备及存储介质
CN109325972B (zh) * 2018-07-25 2020-10-27 深圳市商汤科技有限公司 激光雷达稀疏深度图的处理方法、装置、设备及介质
CN109447911B (zh) * 2018-10-18 2020-06-09 百度在线网络技术（北京）有限公司 图像复原的方法、装置、存储介质和终端设备
CN111523349A (zh) * 2019-02-01 2020-08-11 珠海金山办公软件有限公司 一种纠偏处理的方法、装置、计算机存储介质及终端
CN110136828A (zh) * 2019-05-16 2019-08-16 杭州健培科技有限公司 一种基于深度学习实现医学影像多任务辅助诊断的方法
CN110516665A (zh) * 2019-08-23 2019-11-29 上海眼控科技股份有限公司 识别图像叠加文字区域的神经网络模型构建方法与系统
DE102019214402A1 (de) * 2019-09-20 2021-03-25 Robert Bosch Gmbh Verfahren und vorrichtung zum verarbeiten von daten mittels eines neuronalen konvolutionsnetzwerks
CN111223065B (zh) * 2020-01-13 2023-08-01 中国科学院重庆绿色智能技术研究院 图像矫正方法、不规则文本识别方法、装置、存储介质和设备
CN111353494B (zh) * 2020-04-09 2023-06-23 成都睿琪科技有限责任公司 钢材编码识别系统及方法
CN111583099A (zh) * 2020-04-14 2020-08-25 上海联影智能医疗科技有限公司 图像摆正方法、计算机设备和存储介质
CN111610901B (zh) * 2020-05-11 2021-10-12 上海翎腾智能科技有限公司 一种基于ai视觉下的英语课文辅助教学方法及系统
TWI790471B (zh) 2020-08-26 2023-01-21 財團法人工業技術研究院 基於深度學習的影像校正方法及系統
CN112836076A (zh) * 2021-01-27 2021-05-25 京东方科技集团股份有限公司 一种图像标签生成方法、装置及设备
US20220408098A1 (en) * 2021-06-18 2022-12-22 Tencent America LLC Block-wise entropy coding method in neural image compression
CN113313117B (zh) * 2021-06-25 2023-07-25 北京奇艺世纪科技有限公司 一种识别文本内容的方法及装置
US20230094787A1 (en) * 2021-09-30 2023-03-30 Adobe Inc. Utilizing machine-learning based object detection to improve optical character recognition
EP4242988A1 (fr) * 2022-03-11 2023-09-13 Tata Consultancy Services Limited Procédé et système pour détecter un texte à partir d'un contenu multimédia capturé à une scène
CN116758559B (zh) * 2023-08-15 2023-11-07 山东唐和智能科技有限公司 一种图文文字转换识别系统及方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105809164A (zh) * 2016-03-11 2016-07-27 北京旷视科技有限公司 文字识别方法和装置
CN106295629A (zh) * 2016-07-15 2017-01-04 北京市商汤科技开发有限公司 结构化文本检测方法和系统
CN106778659A (zh) * 2016-12-28 2017-05-31 深圳市捷顺科技实业股份有限公司 一种车牌识别方法及装置
CN108229470A (zh) * 2017-12-22 2018-06-29 北京市商汤科技开发有限公司 文字图像处理方法、装置、设备及存储介质

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3520041B2 (ja) 2000-10-30 2004-04-19 三洋電機株式会社 画像のあおり補正方法、画像のあおり補正装置およびあおり補正用プログラムを記録したコンピュータ読み取り可能な記録媒体
JP3965983B2 (ja) 2001-11-30 2007-08-29 松下電工株式会社 画像処理方法およびその装置
JP4607633B2 (ja) * 2005-03-17 2011-01-05 株式会社リコー 文字方向識別装置、画像形成装置、プログラム、記憶媒体および文字方向識別方法
US8749839B2 (en) 2005-03-24 2014-06-10 Kofax, Inc. Systems and methods of processing scanned data
CN104899571B (zh) * 2015-06-12 2018-12-04 成都数联铭品科技有限公司 一种用于复杂文字识别的随机样本产生方法
CN104966097B (zh) * 2015-06-12 2019-01-18 成都数联铭品科技有限公司 一种基于深度学习的复杂文字识别方法
CN106127204B (zh) * 2016-06-30 2019-08-09 华南理工大学 一种全卷积神经网络的多方向水表读数区域检测算法
CN106203418A (zh) * 2016-07-14 2016-12-07 北京精英智通科技股份有限公司 一种车牌检测的方法及装置
CN106408038A (zh) * 2016-09-09 2017-02-15 华南理工大学 一种基于卷积神经网络模型的旋转汉字识别方法
US10121250B2 (en) * 2016-12-01 2018-11-06 International Business Machines Corporation Image orientation detection

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105809164A (zh) * 2016-03-11 2016-07-27 北京旷视科技有限公司 文字识别方法和装置
CN106295629A (zh) * 2016-07-15 2017-01-04 北京市商汤科技开发有限公司 结构化文本检测方法和系统
CN106778659A (zh) * 2016-12-28 2017-05-31 深圳市捷顺科技实业股份有限公司 一种车牌识别方法及装置
CN108229470A (zh) * 2017-12-22 2018-06-29 北京市商汤科技开发有限公司 文字图像处理方法、装置、设备及存储介质

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113326724A (zh) * 2021-02-07 2021-08-31 海南长光卫星信息技术有限公司 一种遥感影像变化检测方法、装置、设备及可读存储介质
CN113326724B (zh) * 2021-02-07 2024-02-02 海南长光卫星信息技术有限公司 一种遥感影像变化检测方法、装置、设备及可读存储介质

Also Published As

Publication number Publication date
JP2020532001A (ja) 2020-11-05
CN108229470B (zh) 2022-04-01
CN108229470A (zh) 2018-06-29
JP6961802B2 (ja) 2021-11-05
US11275961B2 (en) 2022-03-15
US20200089985A1 (en) 2020-03-19

Similar Documents

Publication Publication Date Title
WO2019119966A1 (fr) Procédé de traitement d'image de texte, dispositif, équipement et support d'informations
US10984295B2 (en) Font recognition using text localization
US10699166B2 (en) Font attributes for font recognition and similarity
US10937166B2 (en) Methods and systems for structured text detection, and non-transitory computer-readable medium
US9697423B1 (en) Identifying the lines of a table
US8634644B2 (en) System and method for identifying pictures in documents
US9824304B2 (en) Determination of font similarity
EP1999688B1 (fr) Conversion d'images numeriques contenant du texte en fichiers bases sur des jetons dans un but de rendu
US10489893B2 (en) Single image rectification
US11704357B2 (en) Shape-based graphics search
US9842251B2 (en) Bulleted lists
US11341605B1 (en) Document rectification via homography recovery using machine learning
CN108182457B (zh) 用于生成信息的方法和装置
US10049268B2 (en) Selective, user-mediated content recognition using mobile devices
US10403040B2 (en) Vector graphics rendering techniques
CN115620315A (zh) 手写文本检测方法、装置、服务器和存储介质
US20220398398A1 (en) Line removal from an image
US10410052B2 (en) Stroke based skeletonizer
CN114627479A (zh) 结构化文本的提取方法、装置、存储介质及计算机设备
CN111814789A (zh) 一种卡号检测的方法、装置、设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18891633

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2020511273

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 05/10/2020)

122 Ep: pct application non-entry in european phase

Ref document number: 18891633

Country of ref document: EP

Kind code of ref document: A1